In this work, we describe a novel protocol for the quality control evaluation of genomic signatures, called sigQC. We define a streamlined approach for evaluating genomic signatures across multiple datasets, facilitating the translation of a genomic signature from pre-clinical to clinical use.

In modern clinical practice, especially within neurology, genomic signatures are becoming increasingly common. These signatures are often derived from in vitro experiments, or from small patient cohorts. Thus, it is crucial to establish whether the expression patterns and statistical properties of a set of genes, or signature, are conserved across cohorts, datasets, and experimental conditions. Moreover, it is increasingly necessary to compare established genomic signatures on newly obtained datasets to better understand how clinically-relevant they remain.

sigQC is an R software package, freely available at https://cran.r-project.org/web/packages/sigQC/. It is accessible to users with basic familiarity in R and a good grasp of core statistical principles. Through sigQC, the user invokes a number of steps through a single line of code, facilitating the evaluation of their genomic signature on a dataset of their choice.

sigQC elicits critical quality control steps for a clinically-useful genomic signature, including evaluating its structure and summary scoring metrics, while also presenting statistical characteristics in an actionable and readily-interpretable manner. We demonstrate this protocol with a pre-clinical genomic signature for metastasis on a patient-derived genomic dataset, and in doing so, highlight the approach of sigQC.

sigQC is a novel protocol that facilitates a standardized and simplified approach to genomic signature quality control. In doing so, it works to translate signatures for diagnostics, therapeutics, and prognostics from bench to bedside, applicable throughout neurology, and more generally, medicine. Already, sigQC has been adopted by groups across the US and UK, as well as an international consortium for colorectal cancer genomic signatures. Full protocol to appear in Nature Protocols.